M
Mistral AI API Pricing
Updated April 2026 · 6 models
Complete pricing for all Mistral AI API models — Magistral (reasoning), Mistral Large, Medium, Small, Codestral, and Nemo. Mistral offers strong open-weight alternatives with European data residency options.
🧠Magistral: reasoning model
💻Codestral: code-specialized
🇪🇺EU data residency available
⚡Mistral Nemo: ultra-cheap
Mistral's reasoning model — strong for complex analytical and math tasks
$2
input /1M
$5
output /1M
Frontier-level MoE model (675B total / 41B active params) at competitive price
$0.5
input /1M
$1.5
output /1M
State-of-the-art performance at 8x lower cost than previous generation
$0.4
input /1M
$2
output /1M
Efficient small model for simple, high-volume tasks
$0.1
input /1M
$0.3
output /1M
Code-specialized model with 256k context — optimized for fill-in-the-middle
$0.3
input /1M
$0.9
output /1M
Compact multilingual model with 128k context — great budget option for EU-compliance workloads
$0.1
input /1M
$0.3
output /1M